Voiceover Script Changes: How to Handle Them After Recording

By /

When the script changes after the voice is recorded, treat it as a new, small job and not as a vague “fix.” Collect all the notes into one list, mark exactly which words changed, give the talent a precise place to find each line, record only those lines as pickups, and check every join on playback. Done this way, a late rewrite costs one short session instead of four rounds of email.

This guide is for podcast producers, post-production teams, mixing engineers and studio owners who receive script changes from clients. It covers how to sort the request, how to communicate it, how to track versions and what to check once the new audio comes back.

Audome
Keep Audio Feedback in One Place
Audome brings audio files, project communication, and time-stamped feedback into one hub for producers and post-production teams.

TL;DR

  • A change to the words after recording is a pickup session for new content, not a correction of the original read. Name it that way from the start.
  • Collect every stakeholder’s notes into one consolidated list before asking talent to record anything.
  • Mark edits with a version number, highlighted new wording, clearly struck removals and a pointer to the exact spot (timecode, line number or file name).
  • Keep every delivered version so an old line can be rebuilt without guessing.
  • After the pickup arrives, listen to both joins and the surrounding context, not only the replaced line.

Table of Contents

First, decide what kind of change it is

The word “revision” covers very different requests, and mixing them up is where most arguments start.

A retake fixes a mistake by the performer: a mispronunciation, a skipped word, a misread line. The script did not change. VO Strategist’s explanation of retakes versus revisions puts it plainly: a retake is the performer’s error, while a revision happens because the client changed the script after the audio was submitted.

A performance revision asks for the same words read differently: slower, warmer, more emphasis. One round is often part of the deal.

A script rewrite changes the words. The Voice Realm explains that when the wording changes after the agreed version was recorded, the actor is being asked to record material that was never part of the original performance. Tiffany Magnuson makes the same distinction: a script change after recording is a pickup session for new content, and knowing which type you face keeps everything organized.

Terminology varies. One performer says “pickup,” another says “retake,” another files it under “revision.” So do not assume the other side shares your vocabulary. Write down in the request which of the three you mean.

Three types of voiceover recording changes

Why the label matters for scope

A default policy published by Voices says that mispronunciations and missing words are fixed at no charge, one round of pacing or tone feedback is included, and one round of rewrites under 10% of the final script is included. Rewrites beyond 10% are quoted on request. That is one marketplace’s reference copy, not a universal rule, but it shows why the category of the change decides who pays for it. Your own terms should say the same thing in your own numbers.

RealVocal Talent lists what generally falls outside a standard revision: major script changes, a different tone or style, a different voice actor, a changed usage, or direction that contradicts the original brief. Those are scope changes, and they normally mean a new quote.

Consolidate the notes before anyone records

The cheapest improvement you can make is to stop passing notes along one at a time.

  1. Gather every stakeholder’s changes first. The Voice Realm advises consolidating feedback from multiple stakeholders before requesting the next recording. Tiffany Magnuson adds that if three people have notes, one compiled round is faster than three separate pickup sessions.
  2. Set an obvious approval point. Tell the client: once you sign off on this version, the script goes to recording. Anything after that is a pickup.
  3. Ask for specific wording, not impressions. RealVocal Talent’s example: replace “make it sound better” with a request to put more emphasis on a named word, and avoid piecemeal requests that cause delays.
  4. Clarify the risky terms once. Brand names, product names, acronyms and numbers cause many late changes. Alison Pentecost’s producer checklist includes verified pronunciations, clarified numbers and abbreviations, confirmed names, and someone with authority available to answer copy questions.

If you manage client feedback on audio as well as on text, our guide to resolving client comments in one edit pass applies the same consolidation idea to mixes.

How to mark script edits so nobody has to guess

Talent should never have to compare two long documents line by line. The Voice Realm recommends a clear filename or version number, highlighted revised wording, and clear marking of removed sentences, and a clear split between text that changed and text shown only for context. It also warns against sending several partially updated documents in separate messages.

A practical markup:

  • Highlight new or changed wording.
  • Strike through deleted sentences, and leave them visible in that version.
  • Grey out the surrounding sentences included only so the talent can hear the lead-in and lead-out.
  • Put the version in the filename and on the first line, for example “Narration_v3”.

Chilton Webb suggests sending a clean diff with one line per touched paragraph, such as the old opening and the new opening of paragraph 14. The client signs off faster when they can see what changed and what did not. He also recommends receiving revisions as a marked-up script with tracked changes rather than a long email describing them.

Do not mark the delivery for the talent

Marking style is personal. C.C. Hogan notes that voice artists should mark up their own scripts in their own way, because they know what each mark means and the process doubles as rehearsal. Jim Powell’s advice for producers is similar: put pacing and inflection direction in parentheses at the head of the scene instead of marking the talent’s copy. Your markup shows what changed. The performer decides how to read it.

Point to the exact spot

A change note is only useful if the talent can find the moment. Pick the reference that fits the delivery format:

Delivery format Best reference Example
One mastered file Timecode range “Re-record the section starting at 2:45 through 3:10”
Many segmented files File name, line number or module title “Line 14” or “Module 3 intro”
Video project Timecode plus the surrounding section “00:01:12, before ‘Today we launch’”
Audiobook or long-form Chapter and paragraph ID plus first words “Ch. 4, para 7, starts ‘The harbor was…’”

These match what Tiffany Magnuson describes: timecode for a mastered file, line numbers or module titles for segmented delivery. VocalCopyCat recommends a stable ID, an approximate timestamp, the filename and revision reviewed, and the first few words of the passage as a human-readable third cue.

A note format that works

A note format that works

Masahiro Kobayashi suggests writing, in this order, the audio position, target words, desired change, reason and desired result. He also says to classify each instruction first, because if pronunciation, speed, tone, pause and copy changes sit in one comment, nobody can tell which take is correct. His point about timecode is worth repeating to a nervous client: it is a shared number so everyone listens to the same place, not a way to assign blame.

A filled-in example:

v3, 01:12.4, “We are excited to announce” Copy change: replace with “Today we launch.” Reason: legal asked for a present-tense claim. Result: same pace and energy as the line before.

Compare that with “Can you make it more engaging?” Magnuson notes that this kind of request needs clarification before anyone can act on it.

Track versions so everyone works from the approved copy

The next failure after a late change is somebody recording or editing against an old script. A few habits prevent it.

For the audio side of this, see our guides to audio version history and audio version control tools and pitfalls.

Record the pickup so it matches the original

A pickup that sounds like a different day in a different room is worse than no change. The aim, in Steve Gouws’s words about documentary work, is for narrators to recreate tone, pacing and microphone character seamlessly.

For the talent and the studio:

For the delivery of pickups, Antoine Bandele’s narrator directive is a useful template. Send only the pickups, as one unbroken file or one file per session, in script order. Speak the timecode before each revised line, read the full line and not half of it, and leave inserting and mastering to the engineer. He also asks narrators not to mark notes as fixed or resolved, only to leave a note that the pickup was done, so engineers, directors and producers can review them.

Name pickup files consistently too. His example convention is a project acronym, a number, “PickUps”, a session number and “Raw”.

What to check after the new audio arrives

Do not just audition the new line. The damage is usually at the edges.

  1. Listen to both joins. VocalCopyCat advises checking duplicated or removed words, changed character assignment and shifted pauses against the approved manuscript.
  2. Play the surrounding context. Kobayashi notes that the ending of the replacement or the entry into the next cut may have changed. Write the checked filename back into the table so unchecked audio is never confused with it.
  3. Compare numbers and names word for word. VocalCopyCat’s example of a good note is concrete: the recording says fifteen, the approved text says fifty.
  4. Check stability across the project. Gouws advises asking whether the voice feels stable across scenes, pacing stays natural and new pickups match earlier material.
  5. Close the note with the final file. Give each note an ID and status, and record which audio revision closed it.

If the replacement still feels wrong, classify the problem: pronunciation, pause, pace, wrong voice or source text error. A single vague complaint sends everyone back into another round.

Our view: a late script change is a tracking problem

We build Audome for people who live this every week: producers, mixing and mastering engineers, podcast teams and post-production facilities who get a “small copy tweak” after the files have been delivered.

In our experience the tweak itself is rarely the problem. The problem is that the feedback arrives in five places, nobody knows which version the client heard, and the same note is answered twice. That is why we designed Audome around one hub: you upload the audio at up to 96kHz/24-bit, collaborators comment on the waveform with time-stamped remarks, and versions are tracked automatically so the approved one is never in doubt.

A timecoded comment is the same thing Kobayashi and Magnuson ask for in plain text: a shared number that points everyone at the same second. We simply put it on the waveform. And because clients can review without logging in, the note actually gets written instead of lingering in someone’s inbox.

We also think scope should be settled before the work starts. Audome lets you set revision limits and bill for paid revisions through Stripe, so a rewrite beyond your agreed allowance becomes a clear, polite line item and not an awkward conversation.

Kreg

Prevent avoidable changes before recording

Most late changes are predictable. Tiffany Magnuson has seen one small pronunciation pickup in one project and four rounds from vague direction in another, and says the difference was almost always how clear the initial information was.

A short pre-recording checklist:

For workflow detail on the audio side, read our guides on client revision management for producers, how audio engineers handle revisions and the podcast delivery and revision workflow. When a project wraps, a project recap summary helps lock files and fees.

Keep script changes, audio and approvals in one place

If your script changes arrive by email, your audio by link and your approvals by text message, every late edit turns into detective work. Audome gives audio professionals one hub for the audio itself: lossless file sharing, time-stamped comments on the waveform, automatic version tracking, password-protected and branded client portals, download controls, and revision limits with billing for paid revisions.

Audome

The Studio plan is $12.50 per month or $150 per year. The Pro plan is $48 per month or $576 per year. A free trial comes first. See the Audome pricing page for what each plan includes, or take a look at Audome to see how the workflow fits your studio.

FAQ

What is the difference between a pickup and a revision?

A pickup is a short re-recording of specific lines, usually to fit changed copy or to fix an error. A revision is any change requested after delivery, whether to the script or to the performance. Terms vary between performers, so state in your request whether the words changed or only the delivery.

Can a client change the script after the voiceover is recorded?

Yes, but it is a new recording task. The talent is recording words that were not in the approved version, so it should follow your agreed change policy. Many policies include one small rewrite round and quote larger ones, for example Voices’ reference copy, which includes rewrites under 10% of the final script.

How do I tell the talent exactly which lines changed?

Send the revised script with a version number, highlight new wording, clearly mark deleted sentences and include surrounding lines for context. Add a timecode, line number or file name for each change. A one-line note per touched paragraph removes most questions.

How do I make a pickup sound like the original take?

Keep the same microphone, room and processing, and give the talent the approved take to listen to first. Use a short pre-roll so they enter at the same pace and energy. Then check both joins and the surrounding lines on playback.

Sources

Scroll to Top